Whether contemporary or historical in scope, research on minority-language shift and death has tended to focus predominantly on the sociolinguistic backgrounds for language shift: it asks what causes an unbalance in the distribution of languages across domains, how language attitudes are affected by social developments and how these attitudes influence a population’s decision to maintain their language or not, what the roles of adults and children are in language shift, etc. An often underlying question is how our knowledge about the causes of language shift can prevent that same phenomenon, or even reverse it (Fishman 1991).
Much structural-linguistic work on language shift has been on the effect of shift on the abandoned language; a prime example of this is Dorian’s work (1977, 1981) on language attrition in the speech of semi-speakers of East Sutherland Gaelic. On the other side of the process, attention is given to the structural characteristics of the variety of language that minority-language speakers end up speaking after the language shift. Often these language-shift varieties have substratum influence from the minority language – e.g. Irish features in Hiberno-English (Hickey 2007) or Spanish features in Chicano English (Fought 2003) – and it is exactly these substratum non-standard features that researchers have mostly been interested in.
Even if descriptions often focus on deviations from the standard, one characteristic of language-shift Englishes in fact appears to be a striking similarity to Standard English. Trudgill and Chambers (1991: 2–3) claim that “mainstream dialects”, that is, dialects that although variable are very close to Standard English, are associated “with upper- and middle-class speakers throughout the English-speaking world; with areas out of which Standard English as a social dialect grew historically, i.e. the south-east of England; with most urban areas; with areas which have shifted to English only relatively recently, such as the Scottish Highlands and western Wales; and with recently settled mixed colonial dialect-speaking areas, such as most of North America and Australia” (my italicisation). If such a generalisation about the standard-likeness of language-shift varieties really holds true, this raises the question of why this should be. Filppula (2006: 516) offers an intuitive answer, explaining the relatively greater divergence from Standard English of Hiberno-English compared to Welsh or Hebridean English by two factors: the greater pace of language shift and the lesser role of schooling. Naturalistic language learning and rapid shift are to have led to more substratum influence in Hiberno-English, while Welsh and Hebridean English are more standard-like because of a longer period of bilingualism and the introduction of Standard English through education.
In this paper, I argue that – at least for a particular subset of language-shift varieties of English – the standard-likeness of these varieties is not due to education or to a prolonged period of bilingualism, but is a result of the particular sociolinguistic circumstances and processes in which these varieties arose. My argument is based on data drawn from three cases of language shift that took place in the British Isles roughly from the sixteenth to the nineteenth centuries, viz. in Cornwall, the Isle of Man, and Shetland. These cases are relatively well described with regard to both the social background of the various language shifts and the resulting language-shift varieties, although as is often the case with the historical sociolinguistics of minority languages, even for well-described cases, there may still be a dearth of clear and reliable data. I therefore cannot offer entirely conclusive evidence and may well be overstating my case; however, I will argue that in these specific sociolinguistic situations, and ruling out a role for education, the scenario I sketch is highly plausible and deserves serious consideration alongside traditional accounts of the top-down imposition of Standard English.
I begin this paper with a discussion of important processes in dialect change, detailing similarities and differences on the social and linguistic levels (Section 2). I then discuss two processes through which the target language in language shift may end up standard-like and introduce the concept of apparent standardisation (Section 3). This concept is applied to the three case studies in Section 4. In the conclusion (Section 5), I discuss the relevance of apparent standardisation to ongoing work in social dialectology.
2 Processes in dialect change
In order to be able to differentiate between the results of different processes in dialect change, it is necessary to briefly define them first. The first process is that of standardisation. This term conventionally applies to the four-stage process of development of a standard language described by Haugen (1966): selection of norm, codification of form, elaboration of function, and acceptance by the community. In this paper, however, I use the term standardisation in a much narrower sense and restrict it to a particular type of acquisition of the standard language: a conscious adoption of features that are considered to be part of an agreed-upon standard or norm, a telic movement toward a codified model. This (admittedly unconventional) use of the term maps on to (part of) the fourth stage of Haugen’s model; in order to avoid confusion, I use development of a standard where the full four-stage model is meant.
A clear account of standardisation of spoken language is given by Pedersen (2005) for Scandinavian languages, especially Danish. She links the spread of the standard spoken Danish to both attitudinal factors – prescriptivism and a standard language ideology, propagated in schools (Pedersen 2005: 188) – and to social factors, like urbanisation, industrialisation, and democratisation. In particular, Pedersen (2005: 189) highlights the role of rapidly growing rural towns, which functioned as “the connecting link between […] country and town”. The spread of an urban-based Standard Danish, then, may have taken place following a gravity model, with the standard hopping from town to town before reaching the intervening countryside (Chambers and Trudgill 1980: 197). For Danish dialect speakers to be “influenced by the standard in schools and via modern mass media” (Pedersen 2005: 190) is an example of so-called vertical levelling (Hinskens et al. 2005: 11): the degree of difference between L (low/dialect) and H (high/standard) varieties lessens as the dialect moves towards the standard – and as a result, so do differences between individual L varieties.
This change of dialects towards a spoken standard, also termed dedialectisation, is the last stage of a centuries-long development posited for many European language communities, as charted in detail by Auer (2005). The development of standard written languages, used instead of Latin, allowed for a reading pronunciation of these standards to be targeted as a spoken H variety as well, causing spoken diglossia. Auer (2005: 17) dates the earliest signs of development of a spoken standard in Europe to the fifteenth century, first only among a minority cultural elite, and only from around 1900 among the population at large. Once a spoken standard was available, variable use of standard and non-standard features gave rise to a diaglossic continuum ranging from basilectal traditional dialects to the acrolectal standard. More recently, the social factors also identified by Pedersen have caused traditional dialects to be lost in favour of more widely spoken regiolects, and, in the most advanced cases such as Denmark, the spoken standard as the only surviving spoken variety.
In addition to this gradual development, Auer (2005: 29–30) also recognises that dedialectisation may happen directly from standard/dialect diglossia, without the intervening diaglossia. A prime example of this is the loss of Low German dialects in favour of a standard based on High German dialects. Relevant social factors here, too, are the prestige associated with the standard language, and the complementary loss of prestige for the dialects, which become associated with lower-class and less educated speakers.
2.2 Levelling and koinéisation
Dialect change may also occur in a subconscious, bottom-up (or horizontal), atelic process known as dialect levelling (Hinskens et al. 2005: 11). An important process in dialect levelling appears to be accommodation, a subconscious change in one’s speech towards (the generalised idea one has of) that of one’s interlocutor (Coupland 1984: 49; on the link between accommodation and language change, see Auer and Hinskens 2005). Where there is increased contact between speakers of different dialects, continuous adaptation towards the other may cause permanent changes in the dialects, causing convergence. Dialect levelling is theoretically an atelic process, i.e. it is not the case that one set of dialect speakers adopts another dialect wholesale through accommodation, but if the interactions are demographically or socially asymmetrical, convergence may occur at either side of the midpoint between the dialects in contact. Dialect levelling is thought to be the process behind the loss of highly localised dialect features in the North of England and the emergence of a more general Northern English (e.g. Watt 2002).
In very particular settings, where we have high-intensity contact and no appreciable prior use of the language, we can expect to find koinéisation. 1 This is a composite process in which differences between dialects spoken in a location are minimised over a time of some three generations (Trudgill 2004: 84–89). In addition to the initial mixing of dialects, koinéisation consists of levelling, unmarking (the loss of linguistically marked features), interdialect development (the emergence of new “compromise” features not present in any of the input dialects) and reallocation (the refunctioning of surviving competing variants in the community as allophonic or sociolinguistic variation). If these five processes of koinéisation are followed by focusing, the development of local norms, we speak of new-dialect formation.
Standards, including spoken standards, often arise along the lines of Haugen’s four-stage process outlined above, and spread through (telic) standardisation, but they may also arise and spread in a levelling process. Jones (1994) offers some evidence for this in her study of the formation of an (incipient) standard for spoken Welsh in two speech communities at opposite ends of the country (cf. Section 2.4 below). Standardisation does clearly play a role, as is evident from the rise of spelling pronunciations of the established written standard, but there is evidence for levelling as well in the appearance of North Walian features in the South Walian community, and vice versa. Although standardisation and levelling are presented here as dichotomous, it must be stressed, therefore, that they may apply simultaneously. Supralocalisation can involve telic standardisation-type processes, and dedialectisation can likewise involve atelic levelling-type processes.
2.3 Standardisation and koinéisation
Certain parallels exist between koinéisation and dialect levelling on the one hand, and standardisation and codification (a prerequisite for standardisation) on the other. First, both standard varieties and koinés include features that are shared by a majority of traditional dialects. For standard varieties, this can be linked to the fact that they are often codified to have as wide as possible a basis among the speech community. 2 Koinés include majority features as it is these features that win the numerical contest for adoption by being a target for accommodation in the first stage of new-dialect formation and by subsequently being adopted by second- and third-stage acquirers of the new dialect. 3
Another parallel is that both standard varieties and koinés may appear more simplified than traditional dialects – that is, they often have fewer phonological or morphosyntactic distinctions. For standard varieties, this may have to do with their wide base; koinéised varieties are simplified because accommodating adults (and also child learners, but see Trudgill (2009: 101) on the role of children in complexification), when faced with both merged and unmerged categories, will typically prefer the simpler, merged, system. A final parallel is the reduction of variation in both processes. In the codification of the standard, reduction of variation is usually the explicit goal, while in koinéisation, it is the non-telic result of subsequent generations of learners of a new dialect acquiring only those variants that are in the majority, a process in which minority variants are lost.
A complicating fact is that standards (both written and spoken) are often based on urban vernacular varieties, which may themselves be the outcomes of dialect contact and koinéisation, see e.g. Howell (2006) on Early Modern Dutch urban vernaculars. The fact that koinéisation lies at the basis of standards that may later become the target of standardisation makes it even more difficult to differentiate between the two processes, as the outcomes are likely to look similar; however, from a sociolinguistic perspective, it is important to make the distinction between the processes that generated these outcomes.
Of course, there are differences between koinéisation and standardisation, too. The first of these concerns the social background of the varieties in contact: in cases of koinéisation, these have more or less equal status and power, while in cases of standardisation, there is one variety – the standard – that enjoys a higher status and more power than the other varieties. In this sense, then, koinéisation and standardisation map on to Hinskens et al.’s (2005: 11) distinction between horizontal and vertical levelling. In the cases of language-shift varieties discussed in this paper, the question of status and power becomes recursive: the majority (target) language is more powerful than the minority (abandoned) language, but at the same time, there are power differences between varieties of the majority, and perhaps minority, languages themselves. 4
Many other differences between koinéisation and standardisation derive from this primary social difference. Koinéisation can be characterised as a non-telic process of change that is endogenous to the community in question, while standardisation is a telic and exogenous change. Standardisation-type change, as a process that makes the vernacular more similar to that of a high-prestige group, is likely to be socially stratified; koinéisation, on the other hand, almost by definition affects the entire speech community simultaneously and similarly. Social meaning of variants and social stratification only arise once the koiné has settled or “focused” (Trudgill 1999: 197, Trudgill 2008; see footnote 3).
2.4 Apparent standardisation
The parallels between standardisation and koinéisation make the outcomes of these processes very similar, and therefore the processes may be difficult to distinguish from their outcomes post-hoc. Cf. also Jones (1994: 261):
It could, perhaps, be argued that [dialect levelling] is more drastic in its elimination of regional features for the elevation of one variety usually leaves the others intact, albeit stigmatized, whereas in the ‘heart of the artichoke’ method of standardization [i.e., levelling to a common core] all but one composite variety vanish. However, as far as the nature of the standardized variety itself is concerned, in purely linguistic terms there is very little difference between the end-products of either process.
Jones was fortunate enough to witness the standardisation of modern spoken Welsh in real time, and to be able to see both processes at work – even if she does not identify the conditions under which standardisation or koinéisation is preferred. Bearing in mind the uniformitarian principle (Labov 1994: 21), we must assume that both roads to standard-like language were available in the case of the Early Modern language-shift Englishes discussed in this paper as well. The social and linguistic data that is available for these varieties is considerably less detailed and reliable than Jones’ data for Welsh, but I will argue that in these cases, the evidence points to the development of a standard-like variety through koinéisation and levelling rather than through a targeted adoption of Standard English. Instead of standardisation, then, I argue that in these varieties we find apparent standardisation: the outcome is similar, but we have no evidence of standardisation having occurred. The term apparent standardisation is loosely based on Trudgill’s (2004: 109–112) apparent levelling; in this process, proposed to occur in the second generation of new-dialect formation, we find a loss of demographically minority variants as we would in levelling, but there is no evidence of accommodation.
3 Language shift and the acquisition of the target language
If we consider broadly the various types of majority-language input that a population of shifters might receive in the acquisition of a dominant language, three main types can be distinguished, each with their own consequences for the language-shift variety:
Learners can be presented with invariable standard input. Acquiring this type of input will result in a fairly standard-like language-shift variety. This scenario is most likely to occur where language acquisition happens in a more formal, educational setting.
Learners can be presented with relatively less variable non-standard input, i.e. a specific local dialect. If they acquire this type of input, the resulting language-shift variety will be much like the local dialect, and very unlike the standard language. This scenario involves naturalistic language acquisition in face-to-face contact with majority-language speakers.
Learners may also be presented with highly variable non-standard input, e.g. in a situation where the majority-language population consists of migrants from various different backgrounds and with different local dialects. This scenario also involves naturalistic language learning, but may result in a more standard-like language-shift variety, as minority language shift and majority-language koinéisation occur simultaneously.
We may also posit a fourth scenario, in which there is a combination of naturalistic language learning and formal acquisition in an educational setting. Although this is probably the dominant paradigm in the present day, I assume that, as education, and especially second-language education, was considerably less widespread in the Early Modern period, we can ignore this scenario for our present purposes. The level of majority-language education among minority-language speakers was very low, and even if a speaker would acquire the majority language in such a setting, they would continue to use the minority language with their family and relations in the minority-language community.
3.1 Formal education and standard language
Formal language education is generally focused on the acquisition of the standard variety of a language. In a present-day setting, this is achieved by presenting learners with near-invariable standard-language input, and by negatively evaluating (either implicitly or explicitly) linguistic behaviours that do not conform to the norm (Wolfram 2014). Although we would require time travel to ever find out what form the English-language input in the Early Modern classroom had, it is highly unlikely that the target was anything other than (an interpretation of) standard English.
Mitchell (2012) discusses a range of English grammars and readers from the seventeenth and eighteenth centuries. It is striking that every text she discusses appears to describe English as a uniform, non-variable entity, although this may of course be ideologically inspired. The only text that acknowledges variation, A General Dictionary of the English Language by the Irishman Thomas Sheridan (1780), puts it in a rather negative light (Mitchell 2012: 132). That “English” for all intents and purposes meant “standard English” can further be gleaned from comments in the front matter to William Kenrick’s A New Dictionary of the English Language (1773), where he advises learners to copy the pronunciation of native speakers, but includes “natives of Scotland and Ireland and the provincials of Wales, Yorkshire and Cornwall” among the learners of the language (Mitchell 2012: 129).
3.2 Standardisation in naturalistic language learning
Contrary to classroom learning, the input in naturalistic language learning is not an idealised standard language, but rather the native-speaker vernacular, including all the non-standard features and variation that may occur in this. The research on part-naturalistic language learning discussed in this section suggests, however, that the variety acquired by learners is more standard-like than the input.
Regan et al. (2009) followed Irish learners of French during their year abroad. In this setting, the learners received naturalistic non-standard input from their native-speaker peers, and they showed increased use of the non-standard morphosyntactic variables investigated compared to their stay-at-home peers. They did not, however, use the features with as high a frequency as native speakers, and they did not acquire the sociolinguistic and stylistic constraints on the variables. Despite the non-standard input, then, these students acquired a French that was more standardised than native-speaker speech.
Similarly, Polish immigrant adolescents in the UK showed lower rates of the non-standard [ɪn] variant of the (ing) variable than their locally-born peers in London and Edinburgh (Schleef et al. 2011), and higher rates of standard [ɪŋ]. 5 Again, the non-native learners failed to acquire the appropriate linguistic and sociolinguistic constraints on the variation from naturalistic input, but overall produced less non-standard-like speech than the native speakers.
A final example of the link between (near-)naturalistic language acquisition and standardisation comes from an experiment with native-speaker and immigrant students in Lille, France (Pooley 2006). Non-native speakers of French failed to distinguish between standard and non-standard variants as native speakers could. Assuming non-native speakers in historical situations had similar problems, Pooley suggests that nineteenth-century migrants may have acquired non-standard variants that they perceived as French, their mix of standard and non-standard forms contributing to convergence between the standard and non-standard varieties.
These are three studies of (near-)naturalistic second-language acquisition leading to standard-like language in very different environments. Although naturalistic acquisition was primary in these cases, education was probably a factor that brought the acquired variety more in line with the standard. But, as is the case with the Irish learners of French or the Polish learners of English, exposure to and awareness of the standard does not necessarily lead to the acquisition of standard-like speech; nor must exposure to non-standard forms lead to the acquisition of non-standard speech. 6 Given that learners failed to acquire sociolinguistic knowledge, they may not have been aware which forms constituted a standard target, and which were non-standard. This means that the full range of standard and non-standard variants in the input must be taken into account if we are to explain the standard-likeness of language-shift varieties.
4 Outcomes of language shift in Early Modern Britain
4.1 Cornish English
4.1.1 Situational sketch
The first case study is that of Cornish English. Cornwall came under English dominance in the ninth century, after which English very gradually expanded westward at the expense of the area’s Cornish, a Celtic language. The legendary last speaker of Cornish died in 1777. 7 Reasons given for the language’s decline include the role of the church, changes in social networks, and cultural change (George 1993: 411–414).
The dominant narrative of Cornish language death emphasises the enforcement of English in church. George (1993: 413) calls this the “prime cause of the decline” of Cornish, and Jenner (1904: 12) claims that “[t]he Reformation did much to kill Cornish”. Repression of Cornish was already commonplace after the (secular) uprising of 1497, but was increased in severity after the Cornish rebelled against the imposition of the English-language Book of Common Prayer in the 1549 Act of Uniformity (Kent 2006: 10; Mills 2010: 196–200). Even if there may have been some room for Cornish in the religious domain (Berresford Ellis 1971: 14), the church was instrumental in ending the tradition of Cornish-language mystery plays in this period (Smith 1947: 3), removing a cultural stronghold for the language. The use of English code-switches in sixteenth-century mystery plays (by devils, demons and tyrants) suggests that English was understood, if negatively evaluated, but as the plays rework earlier material, the form of the language is not necessarily reflective of that in use in Cornwall at the time (Mills 2012).
Equally important as the influence of the church were the changes to Cornish social networks that took place during the sixteenth century. The incorporation of Brittany into France caused Cornish-Breton trade and religious links to be cut, and contact with speakers of the closely related Breton language to be lost. At the same time, contact with speakers of English increased as numerous English migrated to Cornwall to work in the tin mines, and Cornish speakers were employed in the English army and (merchant) navy (Smith 1947: 7–9). There, they will have met speakers of a broad range of traditional English dialects, but may also have become aware of Standard English and of negative evaluations of non-standard English (Kent 2006: 16).
4.1.2 Description and commentary
Cornish English is an excellent example of Trudgill and Chambers’s observation on the standard-likeness of language-shift Englishes. In fact, their observation was preceded by Jespersen (1946: 60), who wrote: “In Cornwall and in the Scilly Islands, where Keltic was spoken till 150 years ago, I was struck by the ‘pure’ English talked by the peasantry, as compared for example with the dialect of the neigbouring [sic] county of Somerset”. Jespersen’s comment needs some qualification, however, as there is a marked difference between the dialects of eastern Cornwall, where Cornish was lost before 1500, and western Cornwall, where the language survived for up to three centuries longer. Wakelin (1975: 22) describes Truro in central Cornwall as “the western limit of dialect in English”, with a traditional dialect similar to that of Devon spoken in eastern Cornwall. In the west, “many of the phonological types present […] have been considered to be old Standard English ones” (Wakelin 1975: 203).
This is confirmed by metalinguistic comments from the sixteenth and seventeenth centuries, presented in Wakelin (1975: 89–94). A few excerpts are especially worth mentioning. William Scawen, c. 1680, wrote that “tis observed also elsewhere in this county further west, where the Cornish hath been most spoken, that the English thereabouts is much better than the same is in Devon, or the places bordering on them, by being most remote from thence from whence the corruption proceeds”. This was followed in 1695 by Bishop William Camden: “Their Language too, is the English; and (which is something surprizing) observ’d by Travellers to be more pure and refin’d than that of their neighbours, Devonshire and Somersetshire. The most probable reason whereof, seems to be this; that English is to them an introduc’d, not an original Language; and those who brought it in were the Gentry and Merchants, who imitated the Dialect of the Court, which is the most nice and accurate”. Some features of the “naughty Englyshe” (i.e. dialect) spoken in eastern Cornwall were given by Borde in 1542: forms like iche cham ‘I am’, dycke ‘thick’, dyn ‘thin’, dryn ‘therein’, vyshe ‘fish’, and volke ‘folk’, showing features that are still a part of traditional eastern Cornish English (Wakelin 1984), are traditionally associated with Devon and Somerset.
4.1.3 The role of education
The educational history in seventeenth- and eighteenth-century Cornwall is overall very poorly described, with only very few references in the literature. Grammar schools had been established in eight Cornish towns by the mid-seventeenth century, and there is very little to suggest that education was within reach for any appreciable proportion of Cornwall’s predominantly rural population (Stephens 1977: 4). By the late-eighteenth century, it would appear that education was more widespread, with “some kind of school”, for example an SPCK charity school, in “most villages and small towns” (Austin 2000: 49, 60).
Evidence of basic literacy in Cornwall supports this view of the development of education. Analysis of signatures in a 1642 document that all adult males were expected to sign shows an average illiteracy rate of 67.7%. There was, of course, variation: illiteracy rates were higher in parishes further West, and the two parishes for which information on female literacy is available show that less than 10% of women could sign their name (Stephens 1977). Of course, literacy is not as black and white as an analysis of signatures versus marks suggests. There will have been many signatories whose literacy skills were stretched to their limits when they labouriously scribbled the characters that made up their names. There are other reasons as well for Stephens (1977: 2) to consider the 67.7% as a minimum rate of illiteracy – or alternatively, to consider 32.3% as an absolute maximum percentage of the adult male population in Cornwall who had enjoyed even the most minimal amount of schooling. By the 1790s, however, writing was widespread and even common practice among the working classes, and a minority had even mastered the rules of standard written English (Austin 2000). 8 Even if models of standard written English were available through church and education, it is unclear how these translated into spoken interaction, and how they may have influenced the language-shift variety of Cornish English.
4.2 Manx English
4.2.1 Situational sketch
The Isle of Man came into English hands in the fourteenth century. This did not have immediate consequences for the island’s Celtic language, Manx, which survived alongside English for centuries until language shift started in the nineteenth century. The supposed “last native speaker” of Manx died in 1974, but as late twentieth-century revitalisation efforts have been somewhat successful, there are still speakers of Manx on the island today (Broderick 1999: 13, 41–44).
There is a general idea that education contributed to the decline of Manx. And sure enough, although both secular and religious education fluctuated between supporting and repressing the language, the overall attitudes towards Manx in the educational system were never very positive. But as an education system was set up in the seventeenth century and Manx did not start to become lost until the nineteenth, changes in Manx social networks in the eighteenth and nineteenth centuries seem at least as important. Manx speakers developed closer contacts with English speakers through trade and smuggling, improved transport connections, and even tourism. There was also a significant immigration of English speakers from the North West of England, which – together with a nineteenth-century emigration of Manx speakers to the United States – increased the proportion of English speakers in the Manx population (Broderick 1999: 23–26).
4.2.2 Description and commentary
Manx English also supports the generalisation on the standard-likeness of language-shift English. In the late nineteenth century, Ellis characterised the dialect as more similar to Standard English than the adjacent dialects in England, although he did note close similarities to the English of Lancashire (Broderick 1997: 124). Barry (1984: 76) compared the answers for Manx English for 125 questions from the Survey of English Dialects to different dialects of England, and found that of these, 31 were shared with northern English dialects generally, while 19 were Standard English forms. It is also interesting to note that Manx English appears to share more with north-eastern dialects than with the geographically closer north-western dialects; in general, however, this analysis confirms Ellis’ characterisation of Manx English as more standard-like than other varieties from the North-West.
4.2.3 The role of education
The role of organised education as a factor in the Anglicisation of the Isle of Man seems to be more substantial than we saw in Cornwall. After the Restoration, an English schooling programme was set up; the Manx were “ripe for the righteous hand of English civilization to be laid upon them” (Clamp 1988: 11). The pivotal figure in this development was Bishop Barrow, according to whose plan schools were set up in all parishes, to be staffed by existing clergy. These clergy were Manx natives, but had received their education from English clerics (albeit locally in the Isle of Man), and as a result were bilingual in Manx Gaelic and English. A grammar school was set up at Castletown, but here the schoolmaster was an Englishman lured to the island by a nice salary (14). Attendance was compulsory, so all children were affected by the English-language schooling provided under Barrow’s scheme. The project was highly successful, and although attitudes to the English schools varied, “the possession of English and English schooling came to be realized as a benefit to the Manx, anxious to escape from the hand of Church and poverty” (Clamp 1988: 20). The English schools on the Isle of Man expanded during the eighteenth century, and especially after the settlement of many English speakers and the repatriation of Manx soldiers, who had fought in the Napoleonic Wars and had acquired English in the army (Clamp 1991). We have no evidence of the variety of English promoted in these schools – this may have been Standard English – so we cannot rule out that formal English-language education contributed to the more standard-like nature of Manx English.
4.3 Shetland Scots
4.3.1 Situational sketch
The Northern Isles, i.e. Orkney and Shetland, came under control of the Scottish throne and parliament in the years 1468–1472 (Crawford 1967–1968, Crawford 1969, Crawford 1983). The islands were previously part of Denmark-Norway, and the inhabitants spoke Norn, a West Scandinavian language most closely related to Faroese. There had been contacts with Scotland long before the political takeover, mostly in the areas of trade and religion, and it is unlikely that English, or rather Scots, was completely unknown in the islands. Still, the evidence suggests that Scots – the language and its speakers – was not widely present in the islands, especially in more remote Shetland, before the last quarter of the fifteenth century (Donaldson 1983: 10).
The incorporation of Shetland into Scotland, however, caused great demographic upheaval in the sixteenth century as Scots migrated to the islands in great numbers. Within a few generations, the proportion of Scots in the Shetland population grew from virtually zero in 1500 to around a third 100 years later (Donaldson 1983). The geographical background of the migrants was diverse, with different areas of especially the East of Scotland represented (this can be deduced from linguistic evidence in Millar 2008: 247; see also Catford 1957a: 57). It is generally believed that alongside political change and despite links with Norway continuing to some extent (Smith 1990), the migration of Scots to Shetland was a major contributing factor to the language shift from Norn to Scots (Millar 2008; Knooihuizen 2009).
There is clear evidence of Scots being used as the language of law, administration and religion by the mid-sixteenth century (Barnes 1991: 446–447, 451), taking over from Latin (Norn was never written). When Norn became extinct as a common vernacular in the islands is a much more contentious issue, though. Some hold that the language continued to be used until the nineteenth century, and without much contact influence from Scots at that (Rendboe 1984; Wiggen 2002), but the majority consensus – based on, in my opinion, the most reliable evidence – is that Norn ceased to be natively acquired not much after 1700, and that there were no more speakers of the language by the third or fourth quarter of the eighteenth century (Barnes 1998; Knooihuizen 2008).
4.3.2 Description and commentary
The English that replaced the Scandinavian language Norn in Orkney and Shetland is more accurately described as Older Scots (van Leyden 2004: 16) with a Norn substratum. This is problematic for our assessment of Orkney and Shetland dialects in terms of standardness, as they are clearly different from Standard English. However, until the sixteenth century at least, Lowland Scots had its own incipient standardisation and did not look to England for linguistic models (Kniezsa 1997: 44). We should therefore compare these dialects to Standard Scots – a more or less unknown entity – and be aware of changing attitudes to Scots and Scottish English that may have influenced the metalinguistic commentary from especially the eighteenth century.
Most of the metalinguistic commentary on Orkney is reported in Marwick (1929: 225–226), on Shetland in Stewart (1964: 165–167) with additional comments by local ministers taken from the Old Statistical Account of Scotland (OSA) from the 1790s (Sinclair 1791–1799). Comments on Orkney tend to focus on the lack of a strong local accent: Wallace, in 1700, writes that “All speak English, after the Scots way, with as good an Accent as any County in the Kingdom”; and Liddell, in 1797 (OSA 19. 498), that “the language spoken here is much the same as in the South of Scotland, and strangers remark, with less of a provincial accent”. There is mention of Norn substratum influence, mostly as “Norwegian accent”, but the general picture for Orkney is one of little dialectal Scots.
Comments on the Shetland dialect are much less clear. Early eighteenth-century writers, like Gifford in 1733, evaluate the dialect in similar terms to the descriptions of Orkney above: “the language now spoken here is English, which they pronounce with a very good accent”. By the end of the century, however, assessment is much more negative, e.g. Edmonston in 1809: “The present language of the islands is certainly English; but good English, although well understood, is rarely spoken. […] [P]ersons versant in the phraseology of the different parishes would find no difficulty in maintaining a conversation which would be altogether unintelligible to an Englishman, or even to a native of the low parts of Scotland”. This suggests that Shetland Scots is not standard-like, although this may be due in part to specific Shetland lexicon, and to non-standard phonetic realisations. Catford (1957a: 73) describes Shetland Scots as a conservative variety of Scots, that has maintained the standard twelve-vowel system mostly lost from other Lowland Scots dialects (Catford 1957b). Catford’s (1957b: 115) comment on the Scots of Galloway, which from context we may assume to apply to Shetland Scots as well, lends further support to my hypothesis:
This suggests that the 12-vowel system is a survival, and that 16th century ‘standard’ Scots may have had such a system. This hypothesis is supported to some extent by its occurrence again in Galloway. In this area Scots replaced Gaelic at about the same period (16th century) as it replaced the Shetland Norn – perhaps a little earlier. It is not improbable that the variety of Scots which replaced Gaelic in Galloway was ‘standard’ Scots, rather than a neighbouring local dialect, just as, at a later date (18th century onwards) it is ‘standard’ English which replaces Gaelic in the Highlands.
4.3.3 The role of education
In Shetland, Wiggen (2002: 63) claims that “after English became the general language of education in 1872, Scots gave way to English faster in Orkney and Shetland than in the rest of Scotland”. It is unclear what the basis of this statement is; remember that the dialects are usually described as Scots. Wiggen’s explanation, though, is interesting. He blames the supposed rapid shift from Scots to English on the lack of rooting of Scots among the Shetland population: “Previously, the local islanders had heard not a locally developed or rooted, but an imported normative variety of Scots, developed on the Mainland” (2002: 63).
How this normative variety was acquired is not clear, but again, formal education does not appear to have played a major role. There is no evidence of formally organised education until the eighteenth century, although this may not mean that there was none (Graham 1998: 17). When education was made available from 1713 – after the primary language shift dated to around 1700 – the itinerant nature of the schools meant that, although schooling was available to many, people were not exposed to it for extended periods of time. The level and quality of education have also been questioned (Graham 1998: 32–37).
5 Conclusion: Early Modern language shift and apparent standardisation
In this paper, I set out to explain the findings by Trudgill and Chambers (1991: 2–3) and Jespersen (1946: 60) that the English of those areas where English is a relatively recent introduction is more standard-like than that of areas where the language has a longer standing tradition. Specifically, I considered evidence from three areas where English began to make its mark in the Early Modern period: Cornwall, the Isle of Man, and Shetland. Although the evidence is at times rather scanty, some clear parallels are to be found. Firstly, the outcome of language shift in each case has been described as relatively standard-like. This does not mean that it is identical to the standard language, or that no regional features occur, but it is a valid observation when comparing the language-shift variety to neighbouring traditional dialects which are much more different from the standard. Secondly, as education was not very widespread, it is unlikely that it could have served effectively as a vehicle for the standard language in language shift in any of these cases, in particular in the earlier part of the period. Thirdly, each case of language shift – and acquisition of English – took place in a context where the input consisted of different varieties, be it through in-migration of English speakers or other forms of contact.
Even if a telic process of standard-language acquisition cannot be completely ruled out – the possible use of standard English in sermons and schools could offer a vehicle for standard-language acquisition, even if other forms of English will have been more frequent – the sociolinguistic backgrounds and the outcomes in these three cases are compatible with a scenario of apparent standardisation, the standard-likeness of the language-shift varieties being a result of dialect contact taking place simultaneously with language shift, rather than the standard having served as a model. A necessary condition for apparent standardisation is large-scale dialect contact, and that condition was certainly met. True standardisation, on the other hand, is connected to a standard-language ideology and to a vehicle such as education or literacy. The ideology may have existed; the vehicles for the ideology were at best insufficient for success, but frequently did not exist at all. A horizontal contact explanation for standard-likeness in these varieties is therefore very plausible, and as I have argued, is a more likely explanation than vertical imposition.
It is important to distinguish between standardisation and apparent standardisation not only in a context of target varieties in language shift, but also in dialect change more generally. In the historical cases discussed in this paper, we were fortunate enough to be able to almost exclude standardisation as a possibility due to the absence of a suitable vehicle like education, but in most modern cases it is likely that both processes are relevant to different degrees, depending on the range of dialects in contact and the strength of the standard. Recent work on the standardisation of spoken Faroese, for example, where there is no strong ideology of a spoken standard, suggests that vertical and horizontal levelling processes are at work simultaneously (Knooihuizen 2014). Crucially, the outcome of the completed processes does not allow us to distinguish which cases of standardisation are real, and which are apparent, giving rise to unbalanced interpretations of linguistic histories. If we want to know what really happens in dialect change, we must remember that even if it walks like a duck, and talks like a duck, it may not necessarily be a duck.
The main idea presented in this article is a spin-off from my doctoral thesis completed at the University of Edinburgh in 2009 with support from an Arts and Humanities Research Council postgraduate research grant (ref. 06/127443). I am grateful to Graeme Trousdale and Peter Trudgill for discussions that helped me shape the foundations for this argument during my Ph.D. studies. Dick Smakman encouraged me to return to the unfinished draft a few years later and offered helpful comments on the new version. Comments from three anonymous reviewers and the JHSL editors have helped me improve the article in the final stages. If anything remains amiss with the argument, I only have myself – and a scanty historical record – to blame.
Auer, Peter. 2005. Europe’s sociolinguistic unity, or: A typology of European dialect/standard constellations. In Nicole Delbecque, Johan van der Auwera & Dirk Geeraerts (eds.), Perspectives on variation: Sociolinguistic, historical, comparative, 7–42. Berlin: Mouton de Gruyter. Google Scholar
Auer, Peter & Frans Hinskens. 2005. The role of interpersonal accommodation in a theory of language change. In Peter Auer, Frans Hinskens & Paul Kerswill (eds.), Dialect change: Convergence and divergence in European languages, 335–357. Cambridge: Cambridge University Press. Google Scholar
Austin, Frances. 2000. Letter writing in a Cornish community in the 1790s. In David Barton & Nigel Hall (eds.), Letter writing as a social practice, 43–61. Amsterdam: Benjamins. Google Scholar
Barnes, Michael P. 1991. Reflections on the structure and demise of Orkney and Shetland Norn. In P. Sture Ureland & George Broderick (eds.), Language contact in the British Isles: Proceedings of the Eighth International Symposium on Language Contact in Europe, Douglas, Isle of Man, 1988, 429–460. Tübingen: Niemeyer. Google Scholar
Barnes, Michael P. 1998. The Norn language of Orkney and Shetland. Lerwick: Shetland Times. Google Scholar
Barry, Michael V. 1984. Manx English. In Peter Trudgill (ed.), Language in the British Isles, 167–177. Cambridge: Cambridge University Press. Google Scholar
Baxter, Gareth J., Richard A. Blythe, William Croft & Alan J. McKane. 2009. Modeling language change: An evaluation of Trudgill’s theory of the emergence of New Zealand English. Language Variation and Change 21(2). 257–296. Google Scholar
Berresford Ellis, P. 1971. The story of the Cornish language. Truro: Tor Mark Press. Google Scholar
Broderick, George. 1997. Manx English: An overview. In Hildegard L. C. Tristram (ed.), The Celtic Englishes, 123–134. Heidelberg: Winter. Google Scholar
Broderick, George. 1999. Language death in the Isle of Man: An investigation into the decline and extinction of Manx Gaelic as a community language in the Isle of Man. Tübingen: Niemeyer. Google Scholar
Catford, J. C. 1957a. Shetland dialect. Shetland Folk Book 3. 71–76. Google Scholar
Catford, J. C. 1957b. Vowel-systems of Scots dialects. Transactions of the Philological Society 56. 107–117. Google Scholar
Chambers, J. K. & Peter Trudgill. 1980. Dialectology. Cambridge: Cambridge University Press. Google Scholar
Clamp, Peter G. 1988. English schooling in the Isle of Man, 1660–1700: The Barrovian design. Journal of Educational Administration and History 20(2). 10–21. Google Scholar
Clamp, Peter G. 1991. A question of education: Class, language and schooling in the Isle of Man, 1800–1833. Journal of Educational Administration and History 23(1). 1–14. Google Scholar
Coupland, Nikolas. 1984. Accommodation at work: Some phonological data and their implications. International Journal of the Sociology of Language 46. 49–70. Google Scholar
Crawford, Barbara E. 1967–1968. The Earldom of Orkney and Lordship of Shetland: A reinterpretation of their pledging to Scotland in 1468–70. Saga-Book of the Viking Society 17. 156–176. Google Scholar
Crawford, Barbara E. 1969. The pawning of Orkney and Shetland: A reconsideration of the events of 1460–9. Scottish Historical Review 48. 35–53. Google Scholar
Crawford, Barbara E. 1983. The pledging of the islands in 1469: The historical background. In Donald J. Withrington (ed.), Shetland and the outside world 1469–1969, 32–48. Oxford: Oxford University Press. Google Scholar
Donaldson, Gordon. 1983. The Scots settlement in Shetland. In Donald J. Withrington (ed.), Shetland and the outside world 1469–1969, 8–19. Oxford: Oxford University Press. Google Scholar
Dorian, Nancy C. 1977. The problem of the semi-speaker in language death. International Journal of the Sociology of Language 12. 23–32. Google Scholar
Dorian, Nancy C. 1981. Language death: The life cycle of a Scottish Gaelic dialect. Philadelphia: University of Pennsylvania Press. Google Scholar
Dorian, Nancy C. 2010. Investigating variation: The effects of social organization and social setting. Oxford: Oxford University Press. Google Scholar
Ferguson, Charles A. 1975. Toward a characterization of English foreigner talk. Anthropological Linguistics 17(1). 1–14. Google Scholar
Filppula, Markku. 2006. The making of Hiberno-English and other “Celtic Englishes”. In Ans van Kemenade & Bettelou Los (eds.), The handbook of the history of English, 507–536. Oxford: Blackwell. Google Scholar
Fishman, Joshua A. 1991. Reversing language shift: Theoretical and empirical foundations of assistance to threatened languages. Clevedon: Multilingual Matters. Google Scholar
Fought, Carmen. 2003. Chicano English in context. Basingstoke: Palgrave Macmillan. Google Scholar
George, Ken. 1993. Cornish. In Martin J. Ball & James Fife (eds.), The Celtic languages, 410–468. London: Routledge. Google Scholar
George, Ken & George Broderick. 1993. The revived languages: Modern Cornish and Modern Manx. In Martin J. Ball & James Fife (eds.), The Celtic languages, 644–663. London: Routledge. Google Scholar
Graham, John J. 1998. A vehement thirst after knowledge: Four centuries of education in Shetland. Lerwick: Shetland Times. Google Scholar
Haugen, Einar. 1966. Language conflict and language planning: The case of Modern Norwegian. Cambridge, MA: Harvard University Press. Google Scholar
Hickey, Raymond. 2007. Irish English: History and present-day forms. Cambridge: Cambridge University Press. Google Scholar
Hinskens, Frans, Peter Auer & Paul Kerswill. 2005. The study of dialect convergence and divergence: Conceptual and methodological considerations. In Peter Auer, Frans Hinskens & Paul Kerswill (eds.), Dialect change: Convergence and divergence in European languages, 1–48. Cambridge: Cambridge University Press. Google Scholar
Holmes, Janet & Paul Kerswill. 2008. Contact is not enough: A response to Trudgill. Language in Society 37(2). 273–277. Google Scholar
Howell, Robert B. 2006. Immigration and koineisation: The formation of Early Modern Dutch urban vernaculars. Transactions of the Philological Society 104(2). 207–227. Google Scholar
Jenner, Henry. 1904. A handbook of the Cornish language chiefly in its latest stages with some account of its history and literature. London: Nutt. Google Scholar
Jespersen, Otto. 1946. Mankind, nation and individual from a linguistic point of view. London: Allen. Google Scholar
Jones, Mari C. 1994. A tale of two dialects: Standardization in modern spoken Welsh. In Mair M. Parry, Winifred V. Davies & R. A. M. Temple (eds.), The changing voices of Europe: Social and political changes and their linguistic repercussions, past, present and future, 243–264. Cardiff: University of Wales Press. Google Scholar
Kent, Alan M. 2006. “Bringin’ the dunkey down from the carn”: Cornu-English in context 1549–2005 – a provisional analysis. In Hildegard L. C. Tristram (ed.), The Celtic Englishes IV: The interface between English and the Celtic languages, 6–33. Potsdam: Universitätsverlag Potsdam. Google Scholar
Kerswill, Paul & Ann Williams. 2000. Creating a new town koine: Children and language change in Milton Keynes. Language in Society 29(1). 65–115. Google Scholar
Kniezsa, Veronika. 1997. The origins of Scots orthography. In Charles Jones (ed.), The Edinburgh history of the Scots language, 24–46. Edinburgh: Edinburgh University Press. Google Scholar
Knooihuizen, Remco. 2008. Fishing for words: The taboo language of Shetland fishermen and the dating of Norn language death. Transactions of the Philological Society 106(1). 100–113. Google Scholar
Knooihuizen, Remco. 2009. Shetland Scots as a new dialect: Phonetic and phonological considerations. English Language and Linguistics 13(3). 483–501. Google Scholar
Knooihuizen, Remco. 2014. Variation in Faroese and the development of a spoken standard: In search of corpus evidence. Nordic Journal of Linguistics 37(1). 87–105. Google Scholar
Labov, William. 1994. Principles of linguistic change. Volume 1: Internal factors. Oxford: Blackwell. Google Scholar
Marwick, Hugh. 1929. The Orkney Norn. London: Oxford University Press. Google Scholar
Millar, Robert McColl. 2008. The origins and development of Shetland dialect in light of dialect contact theories. English World-wide 29(3). 237–267. Google Scholar
Mills, Jon. 2010. Genocide and ethnocide: The suppression of the Cornish language. In John Partridge (ed.), Interfaces in language, 189–206. Newcastle-upon-Tyne: Cambridge Scholars Publishing. Google Scholar
Mills, Jon. 2012. The depiction of tyranny in the Cornish miracle plays. In Liam Mac Amhlaigh & Brian Ó Curnáin (eds.), Ilteangach, ilseiftiúil: Féilscríbhinn in ómós do Nicholas Williams/A festschrift in honour of Nicholas Williams, 139–157. Dublin: Arlen House. Google Scholar
Mitchell, Linda C. 2012. Language and national identity in 17th- and 18th-century England. In Carol Percy & Mary Catherine Davidson (eds.), The languages of nation, 123–140. Bristol: Multilingual Matters. Google Scholar
Pedersen, Inge Lise. 2005. Processes of standardisation in Scandinavia. In Peter Auer, Frans Hinskens & Paul Kerswill (eds.), Dialect change: Convergence and divergence in European languages, 171–195. Cambridge: Cambridge University Press. Google Scholar
Pooley, Tim. 2006. The linguistic assimilation of Flemish immigrants in Lille (1800–1914). Journal of French Language Studies 16(2). 207–233. Google Scholar
Regan, Vera, Martin Howard & Isabelle Lemée. 2009. The acquisition of sociolinguistic competence in a study abroad context. Clevedon: Multilingual Matters. Google Scholar
Rendboe, Laurits. 1984. How “worn-out” or “corrupted” was Shetland Norn in its final stages? NOWELE 3. 53–88. Google Scholar
Schleef, Erik, Miriam Meyerhoff & Lynn Clark. 2011. Teenagers’ acquisition of variation: A comparison of locally-born and migrant teens’ realisation of English (ing) in Edinburgh and London. English World-wide 32(2). 206–236. Google Scholar
Sinclair, J. (ed.). 1791–1799. The [old] Statistical Account of Scotland, drawn up from communications of ministers of the different parishes. Edinburgh: Creech. Google Scholar
Smith, A. S. D. 1947. The story of the Cornish language: Its extinction and revival. Camborne: n.p. Google Scholar
Smith, Brian. 1990. Shetland, Scandinavia, Scotland, 1300–1700: The changing nature of contact. In Grant G. Simpson (ed.), Scotland and Scandinavia 800–1800, 25–37. Edinburgh: Donald. Google Scholar
Smith, Jennifer, Mercedes Durham & Liane Fortune. 2007. “Mam, my trousers is fa’in doon!”: Community, caregiver, and child in the acquisition of variation in a Scottish dialect. Language Variation and Change 19. 63–99. Google Scholar
Stephens, W. B. 1977. Male and female adult illiteracy in 17th-century Cornwall. Journal of Educational Administration and History 9(2). 1–7. Google Scholar
Stewart, John. 1964. Norn in Shetland. Fróðskaparrit 13. 158–175. Google Scholar
Trudgill, Peter. 1999. The chaos before the order: New Zealand English and the second stage of new-dialect formation. In Ernst Håkon Jahr (ed.), Language change: Advances in historical sociolinguistics, 197–207. Berlin, New York: Mouton de Gruyter. Google Scholar
Trudgill, Peter. 2004. New-dialect formation: The inevitability of colonial Englishes. Edinburgh: Edinburgh University Press. Google Scholar
Trudgill, Peter. 2008. Colonial dialect contact in the history of European languages: On the irrelevance of identity to new-dialect formation. Language in Society 37(2). 241–254. Google Scholar
Trudgill, Peter. 2009. Sociolinguistic typology and complexification. In Geoffrey Sampson, David Gil & Peter Trudgill (eds.), Language complexity as an evolving variable, 98–109. Oxford: Oxford University Press. Google Scholar
Trudgill, Peter & J. K. Chambers. 1991. Introduction: English dialect grammar. In Peter Trudgill & J. K. Chambers (eds.), Dialects of English: Studies in grammatical variation, 1–3. London: Longman. Google Scholar
Tuten, Donald N. 2008. Identity formation and accommodation: Sequential and simultaneous relations. Language in Society 37(2). 259–262. Google Scholar
Uther, M., M. A. Knoll & D. Burnham. 2007. Do you speak E-NG-L-I-SH? A comparison of foreigner- and infant-directed speech. Speech Communication 49(1). 2–7. Google Scholar
van Leyden, Klaske. 2004. Prosodic characteristics of Orkney and Shetland dialects: An experimental approach. Utrecht: Landelijke Onderzoekschool Taalwetenschap. Google Scholar
Vikør, Lars S. 1994. Språkplanlegging: Prinsipp og praksis. Oslo: Novus. Google Scholar
Wakelin, Martyn. 1975. Language and history in Cornwall. Leicester: Leicester University Press. Google Scholar
Wakelin, Martyn. 1984. Cornish English. In Peter Trudgill (ed.), Language in the British Isles, 195–198. Cambridge: Cambridge University Press.Google Scholar
Watt, Dominic. 2002. “I don’t speak with a Geordie accent, I speak, like, the Northern accent”: Contact-induced levelling in the Tyneside vowel system. Journal of Sociolinguistics 6(1). 44–63. Google Scholar
Wiggen, Geirr. 2002. Norns død, især skolens rolle: Kommentarer til en disputt om nedgangen for det nordiske språket på Orknøyene og Shetland. Oslo: Novus. Google Scholar
Wolfram, Walt. 2014. Integrating language variation into TESOL: Challenges from English globalization. In Ahmar Mahbook & Leslie Barratt (eds.), Englishes in multilingual contexts: Language variation and education, 15–32. Dordrecht: Springer. Google Scholar
If a population settles in an area where a version of their language is already established, they will most likely converge with the existing “founder population”. Koinéisation is therefore most applicable to colonial varieties such as New Zealand English (Trudgill 2004), although the process is claimed to have occurred in other settings as well. This includes language-shift varieties such as Shetland Scots (Millar 2008; Knooihuizen 2009) as well as settings where the immigrant population is so much more numerous than the founder population that the latter become “swamped”. This happened, for example, in the development of the new dialect of Milton Keynes (Kerswill and Williams 2000).
Note, though, that different ideologies may apply in the development of a standard language that can result in different outcomes (Vikør 1994: 143–185).
Trudgill (2008) argues that feature selection in new-dialect formation is based solely on linguistic and demographic-numerical factors, and that issues of identity are irrelevant. Various discussants of Trudgill’s argument in the same journal volume agree that identity is unlikely to be so important that it can select a certain feature for telic adoption in accommodation, and that the development of an identity linked to particular features cannot precede but is simultaneous with the development of a new variety (e.g. Tuten 2008: 259). But Holmes and Kerswill (2008: 275) remind us that the speakers in the original dialect mixture did have sociolinguistic knowledge that they must have used in accommodating to certain features and not to others; and indeed, computer modelling of Trudgill’s deterministic model suggests that numbers alone are not enough and that social stratification does play a role (Baxter et al. 2009). If this sociolinguistic knowledge marks standard features as desirable, which need not be obviously the case, such features are advantaged in a levelling or koinéisation scenario as well as in standardisation. What this means for cases that involve second-language speakers, who do not fully succeed in acquiring native-like sociolinguistic competence (see Section 3.2), remains an empirical question.
I take it as a sociolinguistic given that, under normal circumstances, there is variation in a language community, and that this variation is (at least partially) socially stratified. Crucially, however, in communities undergoing language shift, the variation in the abandoned (moribund) language, may no longer have social meaning, and there may no longer be power differences in the community that are expressed through language (Dorian 2010).
Unlike Schleef et al. (2011), I include the third variant, Polish-accented [ɪŋk], in the standard, as [ɪŋ] is most likely the target here; Polish phonotactics, however, do not allow the velar nasal without a following stop.
Part of the naturalistic input may have consisted of “outsider-directed” speech, which is often simplified (Ferguson 1975) and may take the form of more standard-like language. Uther et al. (2007: 5) emphasize the similarity between foreigner-directed and child-directed speech, in that they are directed at audiences with “similar linguistic needs”, i.e. in need of language instruction. Smith et al. (2007) have shown that in bidialectal communities, explicit teaching of children predominantly occurs in Standard English rather than a local dialect. The audience design and linguistic ideology behind this code choice were presumably similar for outsider-directed speech in the Early Modern period.
Cornish was artificially revived in the twentieth century, and the number of (bilingual) speakers of Revived Cornish has grown to around one thousand (George and Broderick 1993: 653).
Note that Austin’s (2000) analysis centres on the town of Bodmin in the east of Cornwall, and that the supposedly standard-like language-shift variety is that of the west.